INDEX

Explanations

structured data formats and their components

oai_token-act-pair · gpt-4o-mini Triggered by @bot

"type" and related capitalization

np_acts-logits-general · gemini-2.0-flash

type="ref"

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

SequentialGroup

-0.60

Handlung

-0.52

jekk

-0.46

car

-0.46

ảnh

-0.45

 Robson

-0.45

IGraphics

-0.44

 Bals

-0.43

 delan

-0.43

she

-0.42

POSITIVE LOGITS

 type

1.28

type

1.24

Type

1.06

getType

1.02

 TYPE

0.99

 Type

0.98

TYPE

0.96

ype

0.94

setType

0.93

YPE

0.90

Activations Density 0.069%

structured data formats and their components

"type" and related capitalization

type="ref"

No Comments

No Known Activations